//flex table opened by JP

Click to See Complete Forum and Search --> : KT133a, memory settings, and benchmarks


plug-n-pray
01-22-2002, 02:32 PM
I noticed that FSB at 133MHz and memory at 133MHz has a higher memory bandwidth than FSB at 115MHz and memory at 148MHz. I guess the FSB speed has a big impact on the measured bandwidth. Is this actually true, or just for benchmark software?

Jimstep
01-23-2002, 07:57 AM
The FSB is the speed the motherboard will run the memory at. If you have good memory, you can push the fsb higher. I have run my memory at 146Mhz. Of course, this speeds everything else up on the motherboard so you have to be careful.

plug-n-pray
01-23-2002, 12:14 PM
Yeah, but what I mean is that memory speed has less impact on memory bandwidth than FSB. If you hold the memory speed constant and increase the FSB, you will see a much bigger increase in memory bandwidth than if you hold FSB constant and increase the memory speed.

Jimstep
01-23-2002, 08:07 PM
Okay...your confusing with the terms "memory bandwidth" and "memory speed". Can you give me an example with numbers?

plug-n-pray
01-24-2002, 02:25 AM
Oh, never mind. I just downloaded the new sandra and I'm getting completely different results, which coincide with my expectations. Maybe there was a bug in my other version of sandra.

Thanks for your reply.